Soil Salinity Inversion in Coastal Corn Planting Areas by the Satellite-UAV-Ground Integration Approach

remote sensing

Article Soil Salinity Inversion in Coastal Corn Planting Areas by the Satellite-UAV-Ground Integration Approach

Guanghui Qi 1,2, Chunyan Chang 2, Wei Yang 3, Peng Gao 2 and Gengxing Zhao 2,*

1 College of Information Science and Engineering, Shandong Agricultural University, Tai’an 271018, China; [email protected] 2 National Engineering Laboratory for Efﬁcient Utilization of Soil and Fertilizer Resources, College of Resources and Environment, Shandong Agricultural University, Tai’an 271018, China; [email protected] (C.C.); [email protected] (P.G.) 3 The Key Laboratory for Quality Improvement of Agricultural Products of Zhejiang Province, College of Advanced Agricultural Sciences, Zhejiang A&F University, Lin’an, Hangzhou 311300, China; [email protected] * Correspondence: [email protected]; Tel.: +86-05388243939

Abstract: Soil salinization is a significant factor affecting corn growth in coastal areas. How to use multi-source remote sensing data to achieve the target of rapid, efficient and accurate soil salinity monitoring in a large area is worth further study. In this research, using Kenli District of the Yellow River Delta as study area, the inversion of soil salinity in a corn planting area was carried out based on the integration of ground imaging hyperspectral, unmanned aerial vehicles (UAV) multispectral and Sentinel-2A satellite multispectral images. The UAV and ground images were fused, and the partial least squares inversion model was constructed by the fused UAV image. Then, inversion model was scaled up to the satellite by the TsHARP method, and finally, the accuracy of the satellite- Citation: Qi, G.; Chang, C.; Yang, W.; UAV-ground inversion model and results was verified. The results show that the band fusion of Gao, P.; Zhao, G. Soil Salinity UAV and ground images effectively enrich the spectral information of the UAV image. The accuracy Inversion in Coastal Corn Planting of the inversion model constructed based on the fused UAV images was improved. The inversion Areas by the Satellite-UAV-Ground results of soil salinity based on the integration of satellite-UAV-ground were highly consistent with Integration Approach. Remote Sens. the measured soil salinity (R2 = 0.716 and RMSE = 0.727), and the inversion model had excellent 2021, 13, 3100. https://doi.org/ universal applicability. This research integrated the advantages of multi-source data to establish a 10.3390/rs13163100 unified satellite-UAV-ground model, which improved the ability of large-scale remote sensing data

Academic Editors: Ephrem to ﬁnely indicate soil salinity. Habyarimana and Nicolas Greggio Keywords: Sentinel-2A; UAV; ground imaging hyperspectral; multi-source remote sensing data; Received: 30 June 2021 soil salinity Accepted: 3 August 2021 Published: 5 August 2021

Publisher’s Note: MDPI stays neutral 1. Introduction with regard to jurisdictional claims in Globally, declining soil quality poses a significant challenge to improving agricultural published maps and institutional affil- productivity, economic growth and a healthy environment [1,2]. In coastal areas, soil iations. salinity and alkalinity are major soil limiting factors for agricultural and land degradation [3]. Soil salinization not only reduces soil quality and land productivity, leading to a decline in crop yield, but also threatens ecological security and sustainable land use [4–6]. With the change of natural environment and the disturbance of human behavior, regional Copyright: © 2021 by the authors. salinization becomes more and more serious, which affects the sustainable development Licensee MDPI, Basel, Switzerland. of agriculture coastal areas to a great extent [7]. It is of great significance for agricultural This article is an open access article production and sustainable development to accurately extract the soil salinization status distributed under the terms and and grasp its spatial distribution law in the main crop corn planting area of coastal areas. conditions of the Creative Commons To improve the accuracy and efficiency of obtaining regional soil salinization spa- Attribution (CC BY) license (https:// tial distribution information is a prerequisite for rational management and utilization creativecommons.org/licenses/by/ of salinized soil [8]. The traditional method of obtaining soil salinity information in the 4.0/).

Remote Sens. 2021, 13, 3100. https://doi.org/10.3390/rs13163100 https://www.mdpi.com/journal/remotesensing Remote Sens. 2021, 13, 3100 2 of 17

corn planting area is mainly through field survey sampling and laboratory testing, which is time consuming and laborious. Remote sensing technology has become a frequently used method for the quantitative analysis of soil salinization information because of its advantages of fast, large-scale, and non-destructive acquisition of ground feature information [9]. At present, ground, UAV and satellite remote sensing technologies provide a powerful means for monitoring the salt content of surface soils and have been widely used in the monitoring of farmland soil salinization [10–13]. Mohammad et al. [14] implemented the monitoring of soil salinity in a large area of Qom County, Qom Province, Iran, based on Sentinel-2A data. Wei et al. [15] took advantage of UAV equipped with Micro-MCA multispectral sensors to obtain images and realized the estimation of soil salinity in a small area of Hetao Irrigation District. Wang et al. [16] utilized a portable spectrometer ASD to obtain soil hyperspectral to construct a soil salinity inversion model in Baidunzi Basin, China, and achieved high model accuracy. However, most satellite images have low spatial resolution, so it is hard to achieve high precision real-time monitoring; UAV technology can obtain images with high temporal and spatial resolution, but the observation range is small, and it is unable to realize large scale monitoring; hyperspectral data can build high precision inversion models, but point information is incapable of monitoring soil salinity in a continuous spatial range. Therefore, due to the mutual constraints between the spatial resolution, spectral resolution and imaging width of the sensor, it is difficult for a single sensor to meet the requirements of large-scale, high precision and rapid soil salinity monitoring simultaneously [17]. Making full use of the complementary advantages of multi-source remote sensing data to carry out satellite-UAV-ground integrated inversion is an possible way to improve the accuracy of remote sensing inversion of regional soil salinity [18–20]. At present, Satellite-UAV-ground multi-source optical remote sensing data fusion has been applied to regional soil salinity inversion [21–25]. Solmaz et al. [26] constructed an inversion model through Landsat 8 and ASTER image fusion to obtain the spatial distribution of soil salinity in the Balikhli-Chay watershed, which improved the accuracy of soil salinity monitoring in the watershed. However, by the affection of inconsistency of sensor bands and satellite data acquisition time, the accuracy of soil salinity inversion after data fusion still needed to be improved. Zhang et al. [27] took advantage of band correction coefficients to normalize the reflectivity of satellite images based on the correlation between UAV and satellite image reflectivity, which improved the accuracy of soil salinity inversion, but the relationship between UAV and satellite image band spectrum information is nonlinear, and it was unable to accurately express the relationship between the two using only normalized coefficients. Sun et al. [28] used the measured soil hyperspectral data and Landsat-8 OLI multispectral data fusion to improve the retrieval accuracy of soil salt, and analyzed the differences of salt remote sensing in different seasons. Jia et al. [29] built a soil salinization estimation model based on the fusion of ASD hyperspectral and Landsat-8 OLI images, which expanded hyperspectral data from isolated point information to pixel and regional scale; however, due to the large spectral differences, it was difficult to effectively establish the corresponding relationship between the two samples, limiting the improvement of the inversion accuracy. In conclusion, the satellite-UAV-ground integration of regional soil salinity inversion is subject to the following limitations at present. First of all, the discrete hyperspectral observation data cannot accurately match the continuous spatial scale of UAV and satellite data; second, the simple linear method is used to fuse remote sensing data of different scales, and the accuracy of the inversion model built therefrom is limited; third, until now most of the inversion research based on satellite-UAV-ground multi-source remote sensing data fusion is based on the data fusion of two platforms, and the research of soil salinity inversion systems based on the integration of remote sensing data of three platforms still needs further exploration. The overall objective of this research is to to explore the nonlinear fusion method of ground imaging hyperspectral and UAV multi-spectral images, build a soil salinity inversion model based on the fused UAV images, and then to try to construct a high Remote Sens. 2021, 13, 3100 3 of 18

the research of soil salinity inversion systems based on the integration of remote sensing data of three platforms still needs further exploration. The overall objective of this research is to to explore the nonlinear fusion method of ground imaging hyperspectral and UAV multi-spectral images, build a soil salinity inversion model based on the fused UAV images, and then to try to construct a high accuracy inversion model based on satellite image through the upscaling method, so as to realize satellite-UAV-ground integrated soil salinity monitoring in a coastal corn planting area.

2. Materials and Methods 2.1. Study Area

Remote Sens. 2021, 13, 3100 The study area is located in Kenli District, which is located at the estuary3 of 17 of the Yellow River Delta, China (37°24′–38°10′N, 118°15′–119°19′E) (Figure 1). The area has a warm temperate continental monsoon climate with four distinct seasons [30]. The terrain in the area is typical of the delta landform, with a slight fan shape from southwest to accuracy inversion model based on satellite image through the upscaling method, so northeast. The soil is developed on the alluvium of the Yellow River, with sandy loam in as to realize satellite-UAV-ground integrated soil salinity monitoring in a coastal corn theplanting majority; area. due to the lateral infiltration of the Yellow River and the jacking and impregnation of sea water, the land is irrigated backward, and the phenomenon of secondary2. Materials salinization and Methods of the soil is more serious [31]. At present, there is still a large amount2.1. Study of unused Area saline-alkali wasteland, which has huge development and utilization potential.The The study main area is crops located are in Kenli winter District, whea whicht, corn, is located rice atand the estuarycotton, of with the Yellow extensive managementRiver Delta, and China low (37 yield.◦240–38 Corn◦100N, planting 118◦150 –119areas◦19 are0E) mainly (Figure 1distributed). The area hasin the a warm southwest andtemperate in the middle, continental and monsoonthe soil salinization climate with variation four distinct is obvious, seasons [ 30which]. The is terrain an ideal in the area for thisarea research. is typical of the delta landform, with a slight fan shape from southwest to northeast. The soil is developed on the alluvium of the Yellow River, with sandy loam in the majority; Based on the investigation of Kenli District, two test areas, A and B (Figure 1), were due to the lateral inﬁltration of the Yellow River and the jacking and impregnation of sea selectedwater, theto carry land isout irrigated field measurement backward, and theof soil phenomenon salinity, ground of secondary imaging salinization hyperspectral of measurementthe soil is more and serious UAV [ 31flight]. At test. present, Test there area is A still was a largea rectangular amount of area unused of saline-alkali70 m× 80 m, and testwasteland, area B was which a rectangular has huge development area of 60 m and × utilization90 m. Through potential. investigation, The main crops the areplanting time,winter tillage wheat, methods, corn, riceand and fertilization cotton, with conditions extensive in management the test areas and A lowand yield.B were Corn basically theplanting same, the areas growth are mainly of corn distributed in the test in areas the southwest was significantly and in the different, middle,and and the the soil soil salt contentsalinization was distributed variation is in obvious, all grades, which which is an ideal was areatypical for thisand research. representative.

FigureFigure 1. 1. StudyStudy area area and testtest areas areas A A and and B. B.

2.2. DataBased Acquisition on the investigation and Processing of Kenli District, two test areas, A and B (Figure1), were selected to carry out field measurement of soil salinity, ground imaging hyperspectral 2.2.1.measurement Ground Data and Acquisition UAV flight test. and Test Preprocessing area A was a rectangular area of 70 m× 80 m, and testField area Binvestigation was a rectangular in the area study of 60 area m × 90was m. carried Through out investigation, on 14–15 theJuly planting 2020, corn seedlingtime, tillage stage. methods, In order and to ensure fertilization the uniform conditions distribution in the test areas of the A and points, B were 3 sample basically points werethe pre-arranged same, the growth in every of corn 5 inkm the × test5 km areas grid was in the significantly study area, different, and finally and the 37 soil soil salt sample pointscontent in wasthe distributedcorn planting in all area grades, were which colle wascted typical for andregional representative. inversion verification. A ground2.2. Data card Acquisition was placed and Processing every 10 m on the peripheral boundary of test area A and B 2.2.1. Ground Data Acquisition and Preprocessing Field investigation in the study area was carried out on 14–15 July 2020, corn seedling stage. In order to ensure the uniform distribution of the points, 3 sample points were pre-arranged in every 5 km × 5 km grid in the study area, and finally 37 soil sample points in the corn planting area were collected for regional inversion verification. A ground card was placed every 10 m on the peripheral boundary of test area A and B respectively, and a 10 m× 10 m sample grid was formed after connecting with the measuring rope. Taking the grid intersection as the sampling point, a total of 142 sample points were collected, including 72 in test area A and 70 in test area B. The 6 outliers in the sampling points were Remote Sens. 2021, 13, 3100 4 of 18

respectively, and a 10 m× 10 m sample grid was formed after connecting with the measuring rope. Taking the grid intersection as the sampling point, a total of 142 sample points were collected, including 72 in test area A and 70 in test area B. The 6 outliers in Remote Sens. 2021, 13, 3100 4 of 17 the sampling points were eliminated, and the remaining 136 samples were used for the construction and verification of the corn soil salinity inversion model. An EC110 portable salinity meter (Spectrum Technologies, Inc., Aurora, CO, USA) equippedeliminated, with and a the2225FST remaining series 136 samplesprobe (con wereductivity used for thetemperature construction correction and veriﬁcation had been of the corn soil salinity inversion model. completed) was used to measure the electrical conductivity of the soil surface 10 cm of An EC110 portable salinity meter (Spectrum Technologies, Inc., Aurora, CO, USA) theequipped corn plant with at aeach 2225FST sample series point, probe and (conductivity the average temperature value of multi-point correction hadmeasurement been wascompleted) taken as wasthe usedEC tovalue measure of each the electrical sample conductivitypoint, in dS/m. of the According soil surface 10to cmthe of previous the researchcorn plant results at each of samplethe laboratory, point, and the the averageformula value SS = of 2.18 multi-point × EC + measurement0.727 was applied was to converttaken the as the measured EC value EC of eachdata sample into the point, soil insalt dS/m. content According (SS), in to g/kg the previous[32]. research resultsHyperspectral of the laboratory, images the were formula obtained SS = 2.18 fr×omEC a + 0.727SOC710VP was applied portable to convert hyperspectral the imagermeasured with EC a spectral data into range the soil of salt 400–1000 content (SS),nm, inspectral g/kg [32 resolution]. of 4.68 nm, 128 bands, and theHyperspectral lens type was images C-Mount were [33]. obtained The fromhyperspectral a SOC710VP image portable acquisition hyperspectral time imager was 11:00– with a spectral range of 400–1000 nm, spectral resolution of 4.68 nm, 128 bands, and the 15:00, the weather was clear and cloudless, and the wind was light. The SOC710VP was lens type was C-Mount [33]. The hyperspectral image acquisition time was 11:00–15:00, the placedweather on wasa tripod, clear and and cloudless, the lens and height the wind was was 1.6 light. m. The Ten SOC710VP hyperspectral was placed images on a were collectedtripod, andin the the four lens corners height was and 1.6 the m. center Ten hyperspectral of the two imagestest areas, were as collected shown inin the Figure four 2a,b. At cornersthe same and time, the center the gray of the plate two testimages areas, were as shown collected in Figure for 2correction.a,b. At the same The time,ground the range of graythe acquired plate images images were collectedwas 60 cm for correction.× 80 cm, Theand groundthe images range were of the acquiredprocessed images to realize reflectancewas 60 cm conversion.× 80 cm, and the images were processed to realize reﬂectance conversion.

FigureFigure 2. UAV 2. UAV images images of of the the test test areas areas (A, B)B) andand ground ground hyperspectral hyperspectral images images (a,b ).(a,b).

2.2.2.2.2.2. UAV UAV Data Data Collection Collection and PreprocessingPreprocessing FromFrom 11:00–15:00 11:00–15:00 on on 14–15 14–15 JulyJuly 2020,2020, the the DJI DJI Matrice Matrice 600 600 Pro Pro six-rotor six-rotor UAV UAV (SZ DJI (SZ DJI TechnologyTechnology Co., Co., Ltd.Ltd. Shenzhen, Shenzhen, Guangdong Guangdon Province,g Province, China) wasChina) equipped was withequipped a Sequoia with a multispectral camera (Parrot Sequoia, Parrot Inc., Paris, France) to obtain UAV images. Sequoia multispectral camera (Parrot Sequoia, Parrot Inc., Paris, France) to obtain UAV The camera can receive a total of 4 bands of information, which are green light (G), red images.light (R), The red camera edge (RE) can and receive near infrareda total of (NIR). 4 bands Its spectral of information, response function which was are showngreen light (G),in redFigure light3 . (R), The Sequoiared edge multi-spectral (RE) and near camera infrared was fixed(NIR). on Its the spectral UAV gimbal, response and thefunction wasSunshine shown sensorin Figure which 3. The measured Sequoia the solarmulti-spectral irradiance wascamera mounted was fixed on the on top the of theUAV UAV, gimbal, andand the the Sunshine radiation sensor correction which data measured were used the to correct solar irradiance the image during was mounted the flight on [34 ].the top of the UAV,The and Sequoia the radiation multispectral correction camera anddata radiationwere used sensor to correct were calibrated, the image and during the the flightground [34]. standard whiteboard image was collected before the UAV took off. The flying height was 50 m, the flying speed was 5 m/s, and the image acquisition interval was 1.5 s. After the data were collected, they were imported into Pix4D Mapper software (Pix4D, Prilly, Switzerland) for splicing, radiation correction and other processing to obtain the high resolution ortho-reflection image of the test area, with a spatial resolution of 5 cm. In order to eliminate the random error caused by the reflectance of a single point, a 5 × 5 pixels-size image was taken with the sampling point as the center, and the average reflectance of the image was taken as the reflectance data of the sampling point. Remote Sens. 2021, 13, 3100 5 of 18 Remote Sens. 2021, 13, 3100 5 of 17

Figure 3. FigureSequoia 3. Sequoia sensor spectral sensor response spectral function. response function. 2.2.3. Sentinel-2A Data Acquisition and Preprocessing The Sentinel-2 satellite comprises two small satellites, A and B, with a revisit period of 5 days.The The mainSequoia payload ismultispectral an MSI multispectral camera imager, covering and theradiation 0.4–2.4 µm sensor were calibrated, and the spectral range, including 10 m (4 bands) and 20 m (6 bands), 60 m (3 bands) ground resolutionground [28 ].standard The Sentinel-2A whiteboard product was downloaded image was from the collec ESA Copernicusted before the UAV took off. The flying dataheight sharing was website 50 (https://scihub.copernicus.eu/ m, the flying speed, accessedwas 5 onm/s, 16 December and the 2020). image acquisition interval was 1.5 s. Considering the acquisition time of ground and UAV data and the quality of the image, the Sentinel-2AAfter the Level-1C data multispectral were collected, image on 15 July they 2020 waswere selected imported for modeling into and Pix4D Mapper software (Pix4D, inversion of soil salinity, and images on 14 August 2020 and 18 October 2020 were selected forPrilly, extraction Switzerland) of the corn planting areas.for splicing, radiation correction and other processing to obtain the highRadiometric resolution calibration, ortho-reflection atmospheric correction image and resampling of the of te L1Cst data area, were with a spatial resolution of 5 cm. In carried out to generate images with a spatial resolution of 10 m, and the Sentinel-2A true colororder image to of the eliminate study area was the obtained random by stitching error and clipping caused (Figure 1by). Four the bands, reflectance of a single point, a 5 × 5 B3,pixels-size B4, B6 and B7, ofimage Sentinel-2A was image taken were selected with which the harmonized sampling with the point UAV as the center, and the average image bands. The central wavelengths are 560 nm, 665 nm, 740 nm and 783 nm respectively, andreflectance the band widths of are the 45 nm, image 38 nm, 18 was nm and taken 28 nm respectively.as the reflectance data of the sampling point.

2.3. UAV-Ground Data Fusion 2.2.3.The mapping Sentinel-2A relationship Data between Acquisition the coincidence area and of the Preprocessing UAV and the ground image was established by utilizing the ground hyperspectrum and UAV pixel reflectivity of the homonymyThe Sentinel-2 points in the coincident satellite area comprises to solve the reconstruction two small of the satellites, spectral A and B, with a revisit period information of the non-coincident area. The fusion of high-fidelity space and spectral informationof 5 days. of images The with main different payload widths was is realized, an MSI so as tomultispectral enrich the spectral imager, covering the 0.4–2.4 μm informationspectral of therange, UAV image including [35]. The spatial 10 resolutionm (4 bands) of the ground and hyperspectral 20 m (6 bands), 60 m (3 bands) ground image was resampled to 5 cm, which was the same as the spatial resolution of the UAV image.resolution According to[28]. the spectral The response Sentinel-2A function of theproduct Sequoia sensor, was the downloaded hyperspectral from the ESA Copernicus data data were fitted into UAV band data by convolution operation. The calculation formula is shownsharing in Equation website (1) [36]: (https://scihub.copernicus.eu/, accessed on 16 December 2020). Considering the acquisition time of ground and UAV data and the quality of the image, R λmax s(λ)ρ(λ)dλ λmin ρUAV = (1) the Sentinel-2A Level-1C Rmultispectralλmax image on 15 July 2020 was selected for modeling λ s(λ)dλ and inversion of soil salinity,min and images on 14 August 2020 and 18 October 2020 were where ρUAV is the simulated UAV band reflectivity, s(λ) is the UAV spectral response function,selectedλmax andforλ minextractionare the upper of and the lower corn limits ofplanting the band, respectively, areas. and ρ(λ) is the groundRadiometric hyperspectral data. calibration, atmospheric correction and resampling of L1C data were The relationship between ground hyperspectral and UAV image spectrum was non- linearcarried [37], so theout quadratic to generate polynomial images model was usedwith to traina spatial the mapping resolution relationship of 10 m, and the Sentinel-2A true ofcolor the coincident image area andof expandthe study the spectral area information was ofobtained the non-coincidence by stitching area, to and clipping (Figure 1). Four bands, B3, B4, B6 and B7, of Sentinel-2A image were selected which harmonized with the UAV image bands. The central wavelengths are 560 nm, 665 nm, 740 nm and 783 nm respectively, and the band widths are 45 nm, 38 nm, 18 nm and 28 nm respectively.

2.3. UAV-Ground Data Fusion The mapping relationship between the coincidence area of the UAV and the ground image was established by utilizing the ground hyperspectrum and UAV pixel reflectivity of the homonymy points in the coincident area to solve the reconstruction of the spectral information of the non-coincident area. The fusion of high-fidelity space and spectral information of images with different widths was realized, so as to enrich the spectral information of the UAV image [35]. The spatial resolution of the ground hyperspectral image was resampled to 5 cm, which was the same as the spatial resolution of the UAV image. According to the spectral response function of the Sequoia sensor, the hyperspectral data were fitted into UAV band data by convolution operation. The calculation formula is shown in Equation (1) [36]:

λ max λ ρ λ λ  λ s( ) ( )d ρ = min UAV λ (1) max λ λ  λ s( )d min Remote Sens. 2021, 13, 3100 6 of 17

realize the construction of a UAV image with rich spectral information and high spatial resolution. The quadratic polynomial model adopted is shown in Equation (2):

0 2 SU = a × SU + b × SU + c (2)

where a and b are spectral index conversion coefficients, c is the conversion residual, SU is 0 the reflectance of unfused UAV image, and S U is the reflectance of fused UAV image. The reflectivity of pixels with homonymous points on the UAV and ground images were analyzed by fitting degree, and the consistency of spectral information between UAV band unfused and fused and the ground hyperspectrum was explored. The closer the fit degree was to 1, the higher the consistency between the two. The mathematical expression of the degree of fit (R2) is shown in Equation (3) [38]:

∑ (y −y )2 2 = − U,i G,i R 1 2 (3) ∑ (yG,i−yG,i)

where yG,i is the reflectance of a certain point in the ground hyperspectrum, yG,i is the average reflectance of the sample points, and yU,i is the reflectance of the homonymous point in the UAV image. i = l, 2, . . . , n.

2.4. Construction of Soil Salinity Inversion Model Based on Fused UAV Images 2.4.1. Screening of UAV Image Vegetation Index The vegetation index can show the characteristics of vegetation and effectively reflect the growth and health of vegetation [39]. When the degree of soil salinization becomes greater, the visible light reflectance of the vegetation increases, and the near-infrared reflectance decreases [40]. In this research, various vegetation indices related to red light and near-infrared were selected, mainly including normalized difference vegetation index (NDVI), normalized difference red edge index (NDRE), optimized soil adjusted vegetation index (OSAVI), difference vegetation index (DVI), normalized green difference vegetation index (GNDVI), and green-red vegetation index (GRVI). The 6 vegetation indexes were calculated by the fused UAV band spectrum, and the formulas are shown in Table1. The correlation coefficient between each vegetation index and soil salinity was calculated, and the variance inflation factor (VIF) between vegetation indexes was calculated by the formula VIF = 1/(1 − r × r)(r was the correlation coefficient between vegetation indexes) [41], excluding the low correlation or VIF > 10, which was the parameter that cannot be diagnosed by collinearity. The sensitive vegetation indexes were selected for soil salinity modeling.

Table 1. Formulas and corresponding citation for vegetation indexes.

NO. Vegetation Index Formula Citation 1 NDVI (NIR − R)/(NIR + R) 2 NDRE (NIR − RE)/(NIR + RE) [34] 3 OSAVI (1 + 0.16) (NIR − R)/(NIR + R + 0.16) 4 DVI NIR − R 5 GNDVI (NIR − G)/(NIR + G) [42] 6 GRVI (G − R)/(G + R) Note: NIR—Near infrared band; R—Red band; RE—Red edge band; G—Green band.

2.4.2. Construction and Veriﬁcation of Inversion Model The salt contents of 136 samples were sorted from small to large, and the modeling set and validation set were sampled in a ratio of 2:1 to ensure that the construction model samples and the validation samples were evenly distributed. A total of 91 samples were selected for modeling and 45 samples were used for validation. Taking the sensitive vegetation indexes as the model input variable, the corn soil salinity inversion model was Remote Sens. 2021, 13, 3100 7 of 17

constructed by the PLS method [43], which was implemented in Matlab 2016B. The accuracy of model modeling and verification were evaluated by the coefficient of determination (R2) and root mean square error (RMSE). R2 measured the fitting degree of the model, and RMSE reflected the deviation between measured value and predicted value. The closer R2 was to 1, the smaller the RMSE, which means the higher the accuracy of the model, the better the effect. The degree of soil salinization is divided into 5 grades according to soil salinity and relevant criteria [44], non-salinization (<1 g/kg), mild salinization (1–2 g/kg), and moderate salinization (2–4 g/kg), severe salinization (4–6 g/kg) and saline soil (>6 g/kg), and the distribution map of soil salinity grade will be obtained.

2.5. Satellite-UAV Upscale Conversion of Soil Salinity Inversion Model 2.5.1. Information Extraction of the Corn Planting Area The planting area of corn in the study area was extracted by the time series features composed of the NDVI of the Sentinel-2A images on 14 August 2020 and 18 October 2020. Based on the investigation and analysis of the vegetation types in the study area, the middle of August was the vigorous growth period of the vegetation, which was the best time to distinguish vegetation from non-vegetation, and October was the season when the corn had been harvested and the rice and cotton were mature, so the corn ﬁeld was straw or bare land, showing lower NDVI vegetation characteristics. From August to October, the decrease in NDVI of corn crops was the largest, while the decrease in NDVI of other crops was less; the NDVI of forest and other green land remained basically unchanged. Therefore, through NDVI time series feature changes, combined with the training sample data, the following decision rules were established:

 < <  0 NDVI10.18 0.19; Corn = NDVI8.14 > 0.44; (4)  NDVI8.14 − NDVI10.18 > 0.52;

where NDVI8.14 was the NDVI image of Kenli District on 14 August 2020; NDVI10.18 was the NDVI image of 18 October 2020; and the corn planting area in the study area was obtained by masking.

2.5.2. Upscale Transformation of Inversion Model The TsHARP method is often used for downscaling conversion of land surface temperature in remote sensing [45]. The method assumes that the relationship between surface temperature and NDVI remains unchanged at each scale, and the scale conversion of land surface temperature is realized by introducing NDVI to construct trend surface. This research improved the TsHARP method and introduced the PLS model constructed by NDVI, DVI and GRVI as the trend surface to achieve the upscaling conversion of the soil salinity inversion model. Firstly, the relationship between soil salinity and trend surface factor at UAV scale was established: S0.05 = F0.05 (B0.05) (5)

S0.05—The soil salt content inverted by the trend surface transfer function at the UAV scale; B0.05—Trend surface factors at the UAV scale, namely vegetation index NDVI, DVI and GRVI; F0.05—Trend surface inversion function, which was also applicable to the inversion between soil salinity and trend surface factors upscaled to the 10 m spatial resolution of the Sentinel- 2A satellite. Considering that the trend surface may be affected by factors such as soil moisture content, it was difficult for the trend surface factor to fully reflect the distribution of soil salinity, which was reflected in the conversion residual ∆S0.05 on the UAV scale:

∆S0.05 = S − S0.05 (6) Remote Sens. 2021, 13, 3100 8 of 17

where S is the measured value of soil salinity. Due to the error caused by the variation of soil spatial scale, the soil salt content S10 after scale conversion should be constituted of the soil salt content, which was calculated by applying the trend surface established on the UAV scale to the Sentinel-2A data, and ∆S10, which was the conversion residual of the UAV scale ∆S0.05 interpolated to the residual on the satellite scale. The calculation formula was S10 = F0.05 (B10) +∆S10, and the satellite scale soil salinity inversion model was obtained.

2.5.3. Satellite-UAV-Ground Soil Salinity Inversion and Accuracy Veriﬁcation The satellite scale soil salinity inversion model was applied to Sentinel-2A image, and the distribution map of soil salinity in the corn planting area in Kenli District was obtained and compared with the measured values of soil salinity. At the same time, the results of satellite-UAV-ground integrated inversion were compared with the results of soil salinity inversion based on the PLS model built by the direct fusion of UAV and satellite. The R2 and RMSE of soil sampling points’ data and inversion results of two methods were calculated, respectively, and comparative analysis and quantitative evaluation were carried out. In addition, in order to verify the universality of the model, the satellite-UAV-ground integrated inversion model was applied to the Sentinel-2A image of the Kenli corn planting area on 19 July 2019, and the spatial distribution of soil salinity was obtained. The measured soil salinity in the corn planting area and the inverted soil salinity were compared and veriﬁed.

3. Results and Analysis 3.1. Results of UAV-Ground Data Fusion Table2 presents the quadratic polynomial fusion model of four bands based on ground hyperspectral images and UAV multispectral images, as well as the fitting degree of the unfused and fused UAV image with the hyperspectral pixel of the homonymy points. The UAV and the hyperspectral bands have a high degree of fit, all greater than 0.6. The fitting degree of the four bands after fusion was significantly improved compared with unfusion, and the fitting degree of green, red, red-edge and near-red bands and hyperspectral bands were improved by 0.132, 0.096, 0.111 and 0.187, respectively. Therefore, the fusion of UAV and ground band can enrich the spectral information of the UAV image effectively.

Table 2. UAV-ground band fusion model and ﬁtting degree.

Fitting Degree R2 Fusion Model Unfusion Fusion Green Y= 3.254x2 − 0.234x + 0.089 0.629 0.761 Red Y= −0.197x2 + 0.194x + 0.068 0.705 0.801 Redg Y= −0.125x2 + 0.238x + 0.194 0.664 0.775 Nir Y= 1.705x2−0.439x + 0.319 0.611 0.798

3.2. Correlation between UAV Vegetation Index and Soil Salinity Table3 shows the correlation coefficient between vegetation index of fused UAV image and measured soil salinity. In the correlation coefficient matrix, vegetation index and soil salinity all had high correlation; the highest correlation coefficient between NDVI and soil salinity was 0.739, the lowest correlation coefficient between NDRE and soil salinity was 0.440. The VIF values of OSAVI, GNVI and DVI were all higher than 10, showing strong multicollinearity. Therefore, three vegetation indexes, NDVI, DVI and GRVI, were selected as the independent variables to construct the model. Remote Sens. 2021, 13, 3100 9 of 17

Table 3. Correlation coefﬁcient between vegetation index and soil salinity.

r SS NDVI NDRE OSAVI DVI GNDVI GRVI SS 1 NDVI −0.739 ** 1 NDRE −0.440 ** 0.330 ** 1 OSAVI −0.708 ** 0.989 ** 0.328 ** 1 DVI −0.715 ** 0.908 ** 0.287 ** 0.959 ** 1 GNDVI −0.728 ** 0.950 ** 0.408 ** 0.950 ** 0.890 ** 1 GRVI −0.625 ** 0.882 ** 0.484 ** 0.874 ** 0.803 ** 0.935 ** 1 Signiﬁcance levels: ** 0.01.

3.3. The Soil Salinity Inversion Model Based on the UAV-ground Fusion Image and the Inversion Result of the Test Area Table4 illustrates the PLS model and its accuracy constructed by the three sensitive vegetation indexes of NDVI, DVI and GRVI before and after UAV fusion. The accuracy of the PLS inversion model based on the fused UAV images is better than the inversion model established without fusion. The R2 of the modeling set was improved by 0.104, and the RMSE was reduced by 0.22. The R2 of the veriﬁcation set was improved by 0.109 and the RMSE decreased by 0.258. As a consequence, the inversion model based on the fused UAV images can improve the prediction ability of soil salinity.

Table 4. UAV-ground unfusion and fusion inversion model and accuracy.

Modeling Set Validation Set Inversion Model R2 RMSE R2 RMSE

UAV-ground unfusion S0.05 = 5.885–1.669 × NDVI − 8.745 × DVI + 1.273 × GRVI 0.639 0.922 0.617 1.071

UAV-ground fusion S0.05 = 7.375–8.683 × NDVI − 3.083 × DVI + 0.211 × GRVI 0.743 0.702 0.726 0.813

For the salinity of the soil samples, the minimum value was 0.91 g/kg, the maximum value was 7.58 g/kg, and the average value was 4.48 g/kg. Soil samples covered all grades of soil salinization. Soil salinization is common in test areas A and B, and the salinity difference of samples is obvious. Figure4 shows the spatial distribution of soil salinity in test areas A and B obtained by inversion of the PLS model constructed based on the fusion of UAV images. Among the 136 samples, the inverted values of 115 samples were consistent with the measured salinity grade on the ground, accounting for 84.5% of the total. The soil salinization in the two test areas were mainly light and moderate salinization, which was basically consistent with the ﬁeld survey results in the test areas. The result indicates that the soil salinity inversion model based on the fused UAV data can obtain better inversion results and provide more powerful inversion ability.

3.4. Satellite-UAV-Ground Integrated Soil Salinity Upscale Inversion Model 3.4.1. Trend Surface Conversion Function The PLS model based on the fused UAV images was used as the trend surface transformation function, that was F0.05 (B0.05) = 7.375–8.683 × NDVI − 3.083 × DVI + 0.211 × GRVI. Figures5 and6 show the comparisons of the secondary trend surfaces of soil salinity obtained from the inversion of trend surface transfer functions and those measured. The results indicate that the spatial distribution trends of soil salinity in the two trend surfaces are basically the same; the trend surface transfer function can well reﬂect the spatial distribution trend of soil salinity in the study area. Remote Sens. 2021, 13, 3100 10 of 18

the 136 samples, the inverted values of 115 samples were consistent with the measured salinity grade on the ground, accounting for 84.5% of the total. The soil salinization in the two test areas were mainly light and moderate salinization, which was basically consistent with the field survey results in the test areas. The result indicates that the soil salinity inversion model based on the fused UAV data can obtain better inversion results and provide more powerful inversion ability.

Remote Sens. 2021, 13, 3100 10 of 18

the 136 samples, the inverted values of 115 samples were consistent with the measured salinity grade on the ground, accounting for 84.5% of the total. The soil salinization in the two test areas were mainly light and moderate salinization, which was basically Remote Sens. 2021, 13, 3100 consistent with the field survey results in the test areas. The result indicates that the soil 10 of 17 salinity inversion model based on the fused UAV data can obtain better inversion results and provide more powerful inversion ability.

Figure 4. Inversion results and ground measurement of soil salinity in test areas A and B.

3.4. Satellite-UAV-Ground Integrated Soil Salinity Upscale Inversion Model 3.4.1. Trend Surface Conversion Function The PLS model based on the fused UAV images was used as the trend surface transformation function, that was F0.05 (B0.05) = 7.375–8.683 × NDVI − 3.083 × DVI + 0.211 × GRVI. Figures 5 and 6 show the comparisons of the secondary trend surfaces of soil salinity obtained from the inversion of trend surface transfer functions and those measured. The results indicate that the spatial distribution trends of soil salinity in the two trend surfaces are basically the same; the trend surface transfer function can well

reflect the spatial distribution trend of soil salinity in the study area. FigureFigure 4. InversionInversion results results and and ground ground measurement measurement of soil salinity of soil salinityin test areas in testA and areas B. A and B.

3.4. Satellite-UAV-Ground Integrated Soil Salinity Upscale Inversion Model 3.4.1. Trend Surface Conversion Function The PLS model based on the fused UAV images was used as the trend surface transformation function, that was F0.05 (B0.05) = 7.375–8.683 × NDVI − 3.083 × DVI + 0.211 × GRVI. Figures 5 and 6 show the comparisons of the secondary trend surfaces of soil salinity obtained from the inversion of trend surface transfer functions and those measured. The results indicate that the spatial distribution trends of soil salinity in the two trend surfaces are basically the same; the trend surface transfer function can well reflect the spatial distribution trend of soil salinity in the study area.

(a) Trend surface of soil salinity inverted (b) Trend surface of measured soil salinity by trend surface function

FigureFigure 5. 5.Soil Soil salinitysalinity trend surface surface of oftest test area area A. A.

3.4.2. Analysis of Upscale Residual Results

The UAV scale conversion residual ∆S0.05 was interpolated and converted to the satellite scale residual ∆S10. Table5 shows the descriptive statistical features of residuals ∆S0.05 and ∆S10. There were some differences between residuals ∆S0.05 and ∆S10 of different scales. The maximum and standard deviations of ∆S10 were all less than ∆S0.05, and the minimum and average values were all larger than ∆S0.05; the distribution of ∆S10 was (a) Trend surface of soilconcentrated salinity inverted and the (b dispersion) Trend surface was of measured low. Figures soil salinity7 and by8 trenddemonstrate surface function secondary trend Remote Sens. 2021, 13, 3100 11 of 18 surfacesFigure of test 5. Soil area salinity A and trend B constructed surface of test by area UAV A. residuals ∆S0.05 and Sentinel-2A residuals ∆S10, respectively. The spatial distribution structure of residual ∆S0.05 and ∆S10 trend surface was different, but the trend was basically the same.

(b) Trend surface of measured soil salinity (a) Trend surface of soil salinity inverted by trend surface function

FigureFigure 6. Soil 6. Soil salinity salinity trend trend surface surface of of test test area area B. B. 3.4.2. Analysis of Upscale Residual Results

The UAV scale conversion residual ΔS0.05 was interpolated and converted to the satellite scale residual ΔS10. Table 5 shows the descriptive statistical features of residuals ΔS0.05 and ΔS10. There were some differences between residuals ΔS0.05 and ΔS10 of different scales. The maximum and standard deviations of ΔS10 were all less than ΔS0.05, and the minimum and average values were all larger than ΔS0.05; the distribution of ΔS10 was concentrated and the dispersion was low. Figures 7 and 8 demonstrate secondary trend surfaces of test area A and B constructed by UAV residuals ΔS0.05 and Sentinel-2A residuals ΔS10, respectively. The spatial distribution structure of residual ΔS0.05 and ΔS10 trend surface was different, but the trend was basically the same. Table 6 presents the correlation coefficients between residual ΔS10 and NDVI1, DVI1 and GRVI1 of Sentinel-2A images. The correlation coefficients were all over 0.6, indicating a high correlation. The PLS model established by the above three vegetation indexes and the residual ΔS10 was ΔS10 = −1.161 + 2.347 × NDVI1 − 4.505 × DVI1 − 0.08 × GRVI1.

Table 5. Statistical characteristics of residuals error ΔS0.05 and ΔS10.

Statistical Indicators Numerical Value Maximum Minimum Average Standard (g/kg) Value Value Value Deviation ΔS0.05 0.972 −0.997 −0.027 0.524 ΔS10 0.795 −0.813 −0.016 0.461

Remote Sens. 2021, 13, 3100 11 of 17

Table 5. Statistical characteristics of residuals error ∆S0.05 and ∆S10.

Statistical Indicators RemoteNumerical Sens. 2021 Value, 13, (g/kg) 3100 12 of 18 Remote Sens. 2021, 13, 3100 Maximum Value Minimum Value Average Value Standard Deviation12 of 18

∆S0.05 0.972 −0.997 −0.027 0.524

∆S10 0.795 −0.813 −0.016 0.461

(a) Trend surface of residual ΔS0.05 (b) Trend surface of residual ΔS10 (a) Trend surface of residual ΔS0.05 (b) Trend surface of residual ΔS10 Figure 7. The trend surface of the residuals in the test area A. FigureFigure 7. 7.The The trend trend surface surface of of the the residuals residuals in in the the test test area area A. A.

(a) Trend surface of residual ΔS0.05 (b) Trend surface of residual ΔS10 (a) Trend surface of residual ΔS0.05 (b) Trend surface of residual ΔS10 FigureFigure 8. 8.The The trend trend surface surface of of the the residuals residuals in in the the test test area area B. B. Figure 8. The trend surface of the residuals in the test area B.

Table 6. Correlation coefficient between vegetation indexes and residual error ΔS10. TableTable 6. 6.Correlation Correlation coefﬁcient coefficient between between vegetation vegetation indexes indexes and and residual residual error error∆ SΔ10S10. . Vegetation Index VegetationVegetation Index Index r NDVI1 DVI1 GRVI1 r rNDVI NDVI1 1 DVIDVI1 1 GRVIGRVI1 ΔS10 0.702 0.657 0.619 10 ΔS∆S 10 0.7020.702 0.657 0.657 0.619 0.619 3.4.3. Satellite-UAV-Ground Integrated Inversion Model 3.4.3. Satellite-UAV-Ground Integrated Inversion Model 3.5. CornThe Planting trend surface Area and transfer Analysis function of the Results was applied of Soil Salinity to the Inversioninversion of soil salinity at The trend surface transfer function was applied to the inversion of soil salinity at sentinel-2AFigure9 showssatellite the scale, corn plantingand F0.05 area (B10 extracted) = 7.375–8.683 by NDVI × NDVI time series1 − 3.083 and × decisionDVI1 + 0.211 rules. × sentinel-2A satellite scale, and F0.05 (B10) = 7.375–8.683 × NDVI1 − 3.083 × DVI1 + 0.211 × TheGRVI different1 was obtained. colors indicate According the inversion to the above results resi ofdual soil analysis salinity inresults, the corn the planting PLS model area of GRVI1 was obtained. According to the above residual analysis results, the PLS model of obtainedthe satellite by using scale theresidual satellite-UAV-ground ΔS10 was ΔS10 = integrated−1.161 + 2.347 inversion × NDVI model.1 − 4.505 Table × 7DVI illustrates1 − 0.08 × the satellite scale residual ΔS10 was ΔS10 = −1.161 + 2.347 × NDVI1 − 4.505 × DVI1 − 0.08 × theGRVI area1. statisticsTherefore, of the different satellite soil scale salinity PLS grades.inversion The model result of shows soil salinity that soil was salinization S10 = F0. 05 GRVI1. Therefore, the satellite scale PLS inversion model of soil salinity was S10 = F0. 05 (B10) + ΔS10 = 6.214–6.336 × NDVI1 − 7.588 × DVI1 + 0.131 × GRVI1; thus, the (B10) + ΔS10 = 6.214–6.336 × NDVI1 − 7.588 × DVI1 + 0.131 × GRVI1; thus, the satellite-UAV-ground integrated soil salinity inversion model was obtained. satellite-UAV-ground integrated soil salinity inversion model was obtained. Remote Sens. 2021, 13, 3100 12 of 17

was common in the whole corn planting area. The non-salinized soil was less distributed; Remote Sens. 2021, 13, 3100 13 of 18 most areas were mild salinized and moderate salinized soil, accounting for 88.36% of the total area, which are mainly distributed in the relatively high topography of the southwest and central regions. The area of severe salinized and saline soil was small, and it was scattered in the corn planting area. The inversion result was consistent with the actual survey3.5. Corn situation Planting in theArea study and Analysis area and of the the distribution Results of Soil trend Salinity of soil Inversion salinity previously studied [22,28], indicating the validity of the model for accurate inversion of soil salinity. Figure 9 shows the corn planting area extracted by NDVI time series and decision 3.6.rules. Veriﬁcation The different of the Accuracy colors of theindicate Inversion the Model inversion results of soil salinity in the corn 3.6.1.planting Accuracy area Veriﬁcationobtained by of Modelusing Inversionthe satellite-UAV-ground Results integrated inversion model. TableThe 7 soilillustrates salinity obtainedthe area fromstatistics the satellite-UAV-ground of different soil salinity integrated grades. inversion The result and the shows that satellite-UAVsoil salinization inversion was atcommon 37 sampling in the sites whole in the co Kenlirn planting corn planting area. areaThe were non-salinized compared soil was withless thedistributed; measured soilmost salinity areas to were establish mild the sali scatternized plot—the and moderate result is shownsalinized in Figure soil, 10accounting. 2 Thefor R88.36%and RMSE of ofthe the total satellite-UAV-ground area, which integratedare mainly inversion distributed and the in measured the relatively value high weretopography 0.716 and of 0.727, the respectively,southwest and suggesting central the region inversions. The results area and of severe the measured salinized value and saline were high consistency. Compared with the satellite-UAV inversion, the value of R2 was soil was small, and it was scattered in the corn planting area. The inversion result was higher by 0.133 and the value of RMSE was lower by 0.122. Therefore, the consistency betweenconsistent the with results the of actual the satellite-UAV-ground survey situation in integrated the study inversion area and and the the distribution measured trend of datasoil wassalinity higher previously than that ofstudied the satellite-UAV [22,28], indicating inversion, the and validity the satellite-UAV-ground of the model for accurate integratedinversion inversion of soil salinity. model had a better accuracy for the large-scale soil salinity inversion.

FigureFigure 9. 9.Inversion Inversion results results based based on satellite-UAV-ground on satellite-UAV-ground integration. integration.

TableTable 7. 7.Statistics Statistics of soilof soil salinity salinity grade grade area inarea corn in area corn (Unit: area %).(Unit: %).

Soil Salinity Level Non Saline Mild Salinization Moderate SalinizationMild SevereModerate Salinization Severe Saline Soil Soil Salinity Level Non Saline Saline Soil Proportion of Salinization Salinization Salinization 6.21 40.18 48.18 5.31 0.12 inversion result Proportion of 6.21 40.18 48.18 5.31 0.12 inversion result 3.6.2. Veriﬁcation of Model Universality 3.6. TheVerification Sentinel-2A of the image Accu ofracy the cornof the planting Inversion area Model on 19 July 2019 and the measured soil salt content at 44 ground points were used to verify the universality of the inversion model of3.6.1. corn Accuracy soil salinity. Verification Figure 11 is of a scatter Model plot Inversion of the inversion Results results of soil salinity in the corn plantingThe soil area salinity and the obtained measured from soil salinity. the satellite-UAV-ground The result was R2 = 0.605, integratedRMSE = inversion 0.719, and indicatingthe satellite-UAV that the inversion inversion result at had 37 good sampling coherence sites to thein measuredthe Kenli soil corn salinity, planting and the area were modelcompared could bewith used the as anmeasured inversion modelsoil salinity of soil salinity to establish in corn seedlingsthe scatter in coastal plot—the areas. result is shown in Figure 10. The R2 and RMSE of the satellite-UAV-ground integrated inversion and the measured value were 0.716 and 0.727, respectively, suggesting the inversion results and the measured value were high consistency. Compared with the satellite-UAV inversion, the value of R2 was higher by 0.133 and the value of RMSE was lower by 0.122. Therefore, the consistency between the results of the satellite-UAV-ground integrated inversion and the measured data was higher than that of the satellite-UAV inversion, and the satellite-UAV-ground integrated inversion model had a better accuracy for the large-scale soil salinity inversion. Remote Sens. 2021, 13, 3100 14 of 18

Remote Sens. 2021, 13, 3100 13 of 17 Remote Sens. 2021, 13, 3100 14 of 18

Figure 10. Scatter plots of measured sample points and soil salinity inversion results by two methods in Kenli District.

3.6.2. Verification of Model Universality The Sentinel-2A image of the corn planting area on 19 July 2019 and the measured soil salt content at 44 ground points were used to verify the universality of the inversion model of corn soil salinity. Figure 11 is a scatter plot of the inversion results of soil salinity in the corn planting area and the measured soil salinity. The result was R2 = 0.605, RMSE = 0.719, indicating that the inversion result had good coherence to the measured FigureFiguresoil salinity, 10. 10.Scatter Scatter plotsand plots of the measured ofmodel measured sample could pointssample be and used points soil salinityas and an inversion soilinversion salinity results modelinversion by two methodsof resultssoil salinity by two in corn inmethodsseedlings Kenli District. in Kenliin coastal District. areas.

3.6.2. Verification of Model Universality The Sentinel-2A image of the corn planting area on 19 July 2019 and the measured soil salt content at 44 ground points were used to verify the universality of the inversion model of corn soil salinity. Figure 11 is a scatter plot of the inversion results of soil salinity in the corn planting area and the measured soil salinity. The result was R2 = 0.605, RMSE = 0.719, indicating that the inversion result had good coherence to the measured soil salinity, and the model could be used as an inversion model of soil salinity in corn seedlings in coastal areas.

FigureFigure 11. 11.Scatter Scatter plots plots of soil of salinity soil salinity inversion inversion results and result measureds and salinity measured of corn insalinity 2019. of corn in 2019.

4. Discussion 4. Discussion (1) In coastal soil salinization areas, soil salinity is the main factor affecting the growth of corn(1) and In other coastal crops, andsoil othersalinization factors such areas, as soil soil texture, salinity fertility is and the nutrients, main factor and affecting the soilgrowth moisture of havecorn relatively and other balanced crops, effects and on cropother growth. factors Therefore, such as the soil difference texture, fertility and ofnutrients, vegetation and indexes soil only moisture considers have the influence relatively of soilbalanced salinity, andeffects the soilon salinitycrop growth. is Therefore, indirectly inverted through the vegetation index, which has been confirmed by previous studiesthe difference [46–49]. The of relationship vegetation between indexes vegetation only co indexnsiders and soil the salinity influence of crops of undersoil salinity, and the differentFiguresoil salinity 11. environmental Scatter is plots indirectly of conditions soil salinity inverted and inversion at different through result timess and the is measured distinct, vegetation sosalinity the established index,of corn inwhich 2019. has been satellite-UAV-groundconfirmed by previous integrated studies inversion [46–49]. model isThe more relationship suitable for the between inversion vegetation of soil index and salinity4.soil Discussion salinity in the corn of seedlingcrops under stage in different coastal areas. enviro Thenmental model was conditions used to invert and corn at soil different times is salinity at the seedling stage in 2019, and the results confirmed the universal applicability distinct,(1) In so coastal the establishedsoil salinization satellite-UAV-ground areas, soil salinity isintegrated the main inversionfactor affecting model the is more ofgrowth the model. of corn and other crops, and other factors such as soil texture, fertility and suitable(2) Francos for the et al. inversion found that theof traditionalsoil salinity field in non-imaging the corn spectralseedling data stage are discrete in coastal areas. The nutrients, and soil moisture have relatively balanced effects on crop growth. Therefore, andmodel easily was affected used by to the invert soil background corn soil [ 50salinity], while theat the ground seedling hyperspectral stage in imaging 2019, and the results the difference of vegetation indexes only considers the influence of soil salinity, and the technologyconfirmed has the the universal characteristics applicability of image and of spectrum the model. integration, which can accurately soil salinity is indirectly inverted through the vegetation index, which has been determine(2) theFrancos spectral et information al. found ofthe that corn the at the traditional sampling point. field The non-imaging fusion of spectral spectral data are informationconfirmed ofby ground previous hyperspectral studies [46–49]. and UAV The image relationship effectively between improves vegetation the ability ofindex and UAVsoildiscrete imagessalinity and to of accurately easily crops affectedunder express different corn by spectralthe enviro soil information. bacnmentalkground conditions Therefore, [50], while the and accuracy atthe different ground of the times hyperspectral is soildistinct,imaging salt inversion so technology the modelestablished which has the wassatellite-UAV-ground characteristics constructed by UAV of image images integrated and fused spectrum withinversion hyperspectral integration,model is more which can imagingsuitableaccurately data for hasthedetermine been inversion significantly the of soilspectral improved. salinity information in the corn ofseedling the corn stage at in the coastal sampling areas. Thepoint. The model was used to invert corn soil salinity at the seedling stage in 2019, and the results confirmed the universal applicability of the model. (2) Francos et al. found that the traditional field non-imaging spectral data are discrete and easily affected by the soil background [50], while the ground hyperspectral imaging technology has the characteristics of image and spectrum integration, which can accurately determine the spectral information of the corn at the sampling point. The Remote Sens. 2021, 13, 3100 14 of 17

(3) The four bands of the UAV image were fused with the ground hyperspectral image, and the degree of fitting was improved. However, the highest degree of fitting was 0.801, which was still different from the spectral information of the ground hyperspectral image. This is consistent with the results of previous studies [51,52]. Firstly, the divergence, proba- bly due to the quadratic polynomial model used in image fusion, fails to fully explore the internal relationship between the data. The second possibility is the uncertainty of remote sensing data, and the band response functions of different sensors are different. When spectral response function is used for UAV band spectral matching, spectral information will be lost. Therefore, the next step of research should adopt more effective deep-level and high-level feature mining methods, such as deep learning methods [53], and improve the matching accuracy of the spectrum and space of the image to be fused to reduce the influence of radiation and spatial characteristics on the fusion accuracy. The data fusion method in this research has conducted a preliminary study on the fusion of images of different widths and provided a reference for ideas and methodology. (4) The focus of this research was the exploration of the integration method of satellite, UAV and ground. Therefore, only the PLS inversion model was constructed. By the foundation of the effectiveness of integrated satellite-UAV-ground inversion in this research, the next step will be to integrate multi-dimensional features such as spectral features, spatial textures, and crop parameters to construct higher-precision models such as machine learning [54,55], and to screen the optimal model to further improve the accuracy of the regional soil salinity inversion. (5) The heterogeneity of the surface space seriously affects the spatial scale conversion. The trend surface analysis method uses the multivariate statistical method to fit the distribution of spatial elements. In this research, the trend surface was used to simulate the spatial change trend of soil salinity, and the residual surface was used to simulate and quantify the spatial variability of the residual. As the scale increases, the variability of the soil salt residual becomes weaker and the standard deviation decreases; this agrees with the multi-scale variation law of soil salinity in the study area [56]. The influence of information loss and variation in the process of spatial scale upscaling is unavoidable [57,58]. The formulation description of scale transformation and the construction of a better continuous model are worthy of exploration. (6) The satellite-UAV-ground integrated approach proposed in this study improves the monitoring accuracy of regional soil salinization. Compared with the previous satellite- UAV upscale inversion method, the inversion quality of the satellite-UAV-ground integration was better [59–61]. Although satellite-UAV upscaling improves the spatial resolution of satellite images, the improvement of inversion accuracy was still limited by the restric- tion of spectral information. Compared with the traditional scaling up method, the spectral information of the original remote sensing data can be improved, and the spatial structure characteristics can be maintained by introducing the high-resolution spectral information fusion and constructing the trend surface to realize the ascending scale conversion. The integrated satellite-UAV-ground method proposed in this research can be used for large-scale inversion of other surface parameters for reference.

5. Conclusions In this research, ground-based hyperspectral imaging and UAV were fused to construct an inversion model of corn soil salinity, and the model was scaled up to Sentinel-2A satellite scale; the satellite-UAV-ground integrated inversion of soil salinity in the coastal corn planting area was realized. The main conclusions are as follows: (1) The fusion of the four bands of the UAV image with the ground hyperspectral improved the degree of ﬁtting with the hyperspectral data. The vegetation indexes based on the UAV band after fusion had a high correlation with soil salinity. According to the correlation coefﬁcient and variance expansion factor, three sensitive vegetation indexes, NDVI, DVI, and GRVI, were selected as independent variables for PLS Remote Sens. 2021, 13, 3100 15 of 17

modeling and R2 = 0.743; thus, the inversion results were coincident with the actual distribution of soil salinity in the test area. (2) The PLS model S0.05 = 7.375–8.683 × NDVI − 3.083 × DVI + 0.211 × GRVI constructed with the fused UAV images was used as the trend surface conversion function, and the PLS model of the residual ∆S10 was constructed as ∆S10 = −1.161 + 2.347 × NDVI1 − 4.505 × DVI1 − 0.08 × GRVI1. Thus, the Sentinel-2A satellite scale PLS inversion model of soil salinity in the coastal corn planting area of S10 = 6.214–6.336 × NDVI1 − 7.588 × DVI1 + 0.131 × GRVI1 was obtained. (3) The actual soil salinity in the corn planting area was used to verify the inversion results of satellite-UAV-ground integration and satellite-UAV ascending scale, and the inversion results of satellite-UAV-ground were better than those of satellite-UAV inversion and had high consistency with the actual salt distribution. The Sentinel-2A image of corn growing area on 19 July 2019 was used to verify the universality of the model; the R2 of soil salt inversion and measured soil salt was 0.605, which indicated that the model had an excellent universality. (4) The distribution of non-salinized soil in the study area was small, and the majority was mild and moderate salinized soil, accounting for 88.36% of the total area, which was concentrated in the southwest and central part of Kenli District, while the distribution of severe salinized soil and salinized soil was small and scattered in the corn planting area. In this research, we proposed a satellite-UAV-ground integrated soil salinity inversion method in the coastal corn planting area, which provides an effective means for quickly and accurately obtaining soil salinity information in the corn area.

Author Contributions: Conceptualization, G.Q. and G.Z.; methodology, G.Q. and G.Z.; software, P.G.; validation, C.C.; formal analysis, G.Q.; investigation, P.G.; resources, W.Y.; data curation, G.Q. and W.Y.; writing—original draft preparation, G.Q.; writing—review and editing, G.Z. and C.C.; visualization, W.Y.; supervision, P.G.; project administration, C.C.; funding acquisition, W.Y. and G.Z. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by National Natural Science Foundation of China (No. 41877003), Major Science and Technology Innovation Project of Shandong Province (No. 2019JZZY010724), Shandong Province “Double First-Class” Award and Subsidy Fund (No. SYL2017XTTD02), Talent Startup Project of Zhejiang A& F University (No. 113-2034020162). Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Hassan, F.; Mohammad, A.; Brandon, H.; Hamid, S.; Asghar, R.; Abolhasan, F.; Thomas, S.; Ruhollah, T. Spatio-temporal dynamic of soil quality in the central Iranian desert modeled with machine learning and digital soil assessment techniques. Ecol. Indic. 2020, 118, 106736. [CrossRef] 2. Abedi, F.; Amirian-Chakan, A.R.; Faraji, M.; Taghizadeh-Mehrjardi, R.; Kerry, R.; Razmjoue, D.; Scholten, T. Salt dome induced soil salinity in southern Iran–prediction and mapping with averaging machine learning models. Land Degrand. Dev. 2020, 32, 1540–1554. [CrossRef] 3. Girmay, G.; Singh, B.; Mitiku, H.; Borresen, T.; Lal, R. Carbon stocks in Ethiopian soils in relation to land use and soil management. Land Degrand. Dev. 2008, 19, 351–367. [CrossRef] 4. Metternicht, G.I.; Zinck, J.A. Remote sensing of soil salinity: Potentials and constraints. Remote Sens. Environ. 2003, 85, 1–20. [CrossRef] 5. Allbed, A.; Kumar, L.; Aldakheel, Y.Y. Assessing soil salinity using soil salinity and vegetation indices derived from IKONOS high-spatial resolution imageries: Applications in a date palm dominated region. Geoderma 2014, 230, 1–8. [CrossRef] 6. Xu, X.; Zhao, G.X.; Gao, P.; Cui, K.; Li, T. Inversion of Soil Salinity in Coastal Winter Wheat Growing Area Based on Sentinel Satellite and Unmanned Aerial Vehicle Multi-Spectrum—A Case Study in Kenli District of the Yellow River Delta. Sci. Agric. Sin. 2020, 53, 5005–5016. Remote Sens. 2021, 13, 3100 16 of 17

7. Maleki, S.; Karimi, A.; Zeraatpisheh, M.; Poozeshi, R.; Feizi, H. Long-term cultivation effects on soil properties variations in different landforms in an arid region of eastern Iran. Catena 2021, 206, 105465. [CrossRef] 8. Abbas, A.; Khan, S.; Hussain, N.; Hanjra, M.A.; Akbar, S. Characterizing soil salinity in irrigated agriculture using a remote sensing approach. Phys. Chem. Earth 2013, 55, 43–52. [CrossRef] 9. Wang, Z.; Zhang, X.; Zhang, F.; Chan, N.; Kung, H.; Liu, S.; Deng, L. Estimation of soil salt content using machine learning techniques based on remote-sensing fractional derivatives, a case study in the Ebinur Lake Wetland National Nature Reserve, Northwest China. Ecol. Indic. 2020, 119, 106869. [CrossRef] 10. Aldabaa, A.; Weindorf, D.C.; Chakraborty, S.; Sharma, A.; Li, B. Combination of proximal and remote sensing methods for rapid soil salinity quantification. Geoderma 2015, 239, 34–46. [CrossRef] 11. El Harti, A.; Lhissou, R.; Chokmani, K.; Ouzemou, J.E.; Hassouna, M.; Bachaoui, E.M.; El Ghmari, A. Spatiotemporal monitoring of soil salinization in irrigated Tadla Plain (Morocco) using satellite spectral indices. Int. J. Appl. Earth Obs. 2016, 50, 64–73. [CrossRef] 12. Zhang, T.T.; Qi, J.G.; Gao, Y.; Ouyang, Z.T.; Zeng, S.L.; Zhao, B. Detecting soil salinity with MODIS time series VI data. Ecol. Indic. 2015, 52, 480–489. [CrossRef] 13. Nurmemet, I.; Ghulam, A.; Tiyip, T.; Elkadiri, R.; Ding, J.L.; Maimaitiyiming, M.; Abliz, A.; Sawut, M.; Zhang, F.; Abliz, A.; et al. Monitoring Soil Salinization in Keriya River Basin, Northwestern China Using Passive Reflective and Active Microwave Remote Sensing Data. Remote Sens. 2015, 7, 8803–8829. [CrossRef] 14. Mohammad, M.T.; Mahdi, H.; Kamran, E. Retrieval of soil salinity from Sentinel-2 multispectral imagery. Eur. J. Remote Sens. 2019, 52, 138–154. 15. Wei, G.; Li, Y.; Zhang, Z.; Chen, Y.; Chen, J.; Yao, Z.; Lao, C.; Chen, H. Estimation of soil salt content by combining UAV-borne multispectral sensor and machine learning algorithms. PeerJ 2020, 8, e9087. [CrossRef][PubMed] 16. Wang, L.; Zhang, B.; Shen, Q.; Yao, Y.; Zhang, S.; Wei, H.; Yao, R.; Zhang, Y. Estimation of Soil Salt and Ion Contents Based on Hyperspectral Remote Sensing Data: A Case Study of Baidunzi Basin, China. Water 2021, 13, 559. [CrossRef] 17. Berra, E.F.; Gaulton, R.; Barr, S. Assessing spring phenology of a temperate woodland: A multiscale comparison of ground, unmanned aerial vehicle and Landsat satellite observations. Remote Sens. Environ. 2019, 223, 229–242. [CrossRef] 18. Liu, X.; Liu, Q.; Wang, Y.J. Remote sensing image fusion based on two-stream fusion network. Inform. Fusion 2020, 55, 1–5. [CrossRef] 19. Wu, W.; Al-Shafie, W.M.; Mhaimeed, A.S.; Ziadat, F.; Nangia, V.; Payne, W.B. Soil salinity mapping by multiscale remote sensing in mesopotamia, Iraq. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2017, 7, 4442–4452. [CrossRef] 20. He, C.Y.; Gao, B.; Huang, Q.X.; Ma, Q.; Dou, Y.Y. Environmental degradation in the urban areas of China: Evidence from multi-source remote sensing data. Remote Sens. Environ. 2017, 193, 65–75. [CrossRef] 21. Xu, C.; Qu, J.J.; Hao, X.; Cosh, M.H.; Prueger, J.H.; Zhu, Z.; Gutenberg, L. Downscaling of surface soil moisture retrieval by combining MODIS/Landsat and in situ measurements. Remote Sens. 2018, 10, 210. [CrossRef] 22. Ma, Y.; Chen, H.; Zhao, G.X.; Wang, Z.R.; Wang, D.Y. Spectral Index Fusion for Salinized Soil Salinity Inversion Using Sentinel-2A and UAV Images in a Coastal Area. IEEE Access 2020, 8, 159595–159608. [CrossRef] 23. Kattenborn, T.; Lopatin, J.; Förster, M.; Braun, A.C.; Fassnacht, F.E. UAV data as alternative to field sampling to map woody invasive species based on combined Sentinel-1 and Sentinel-2 data. Remote Sens. Environ. 2019, 227, 61–73. [CrossRef] 24. Kattenborn, T.; Eichel, J.; Fassnacht, F.E. Convolutional Neural Networks enable efficient, accurate and fine-grained segmentation of plant species and communities from high-resolution UAV imagery. Sci. Rep. 2019, 9, 9–23. [CrossRef][PubMed] 25. Guo, Y.; Shi, Z.; Huang, J.Y.; Zhou, L.Q.; Zhou, Y.; Wang, L. Characterization of field scale soil variability using remotely and proximally sensed data and response surface method. Stoch. Environ. Res. Risk Assess. 2016, 30, 859–869. [CrossRef] 26. Fathololoumi, S.; Vaezi, A.R.; Alavipanah, S.K.; Ghorbani, A.; Saurette, D.; Biswas, A. Improved digital soil mapping with multitemporal remotely sensed satellite data fusion: A case study in Iran. Sci. Total Environ. 2020, 721, 137703. [CrossRef] [PubMed] 27. Zhang, S.M.; Zhao, G.X. A Harmonious Satellite-Unmanned Aerial Vehicle-Ground Measurement Inversion Method for Monitor- ing Salinity in Coastal Saline Soil. Remote Sens. 2019, 11, 1700. [CrossRef] 28. Sun, Y.N.; Li, X.Y.; Shi, H.B.; Cui, J.Q.; Wang, W.G.; Bu, X.Y. Remote Sensing Inversion of Soil Salinity and Seasonal Difference Analysis Based on Multi-source Data Fusion. Trans. Chin. Soc. Agric. Mach. 2020, 51, 169–180. 29. Jia, P.P.; Sun, Y.; Shang, T.H.; Zhang, J.H. Estimation models of soil water-salt based on hyperspectral and Landsat-8 OLI image. Chin. J. Ecol. 2020, 39, 2456–2466. 30. Zhao, Q.Q.; Bai, J.H.; Gao, Y.C.; Zhao, H.X.; Zhang, G.L.; Cui, B.S. Shifts in the soil bacterial community along a salinity gradient in the Yellow River Delta. Land Degrad. Dev. 2020, 31, 2255–2267. [CrossRef] 31. Wen, Y.; Guo, B.; Zang, W.Q.; Ge, D.Z.; Luo, W.; Zhao, H.H. Desertification detection model in Naiman Banner based on the albedo-modified soil adjusted vegetation index feature space using the Landsat8 OLI images. Geomat. Nat. Hazards Risk 2019, 10, 544–558. [CrossRef] 32. Wang, Z.R.; Zhao, G.X.; Gao, M.X.; Chang, C.Y.; Jiang, S.Q.; Jia, J.C.; Li, J. Spatial variability of soil water and salt in summer and soil salt microdomain characteristics in Kenli County, Yellow River Delta. Acta Ecol. Sin. 2016, 36, 1040–1049. 33. Wen, X.; Zhu, X.C.; Yu, R.Y.; Xiong, J.L.; Gao, D.S.; Jiang, Y.M.; Yang, G.J. Visualization of Chlorophyll Content Distribution in Apple Leaves Based on Hyperspectral Imaging Technology. Agric. Sci. China 2019, 10, 783–795. [CrossRef] Remote Sens. 2021, 13, 3100 17 of 17

34. Dong, C.; Zhao, G.X.; Su, B.W.; Chen, X.N.; Zhang, S.M. Research on the Decision Model of Variable Nitrogen Application in Winter Wheat’s Greening Period Based on UAV Multispectral Image. Spectrosc. Spect. Anal. 2019, 39, 3599–3605. 35. He, J.; Li, J.; Yuan, Q.Q.; Li, H.F.; Shen, H.F. Spatial–Spectral Fusion in Different Swath Widths by a Recurrent Expanding Residual Convolutional Neural Network. Remote Sens. 2019, 11, 2203. [CrossRef] 36. Zhang, F.; Tashpolat, T.; Ding, J.L. Simulation and Subsection between Fields Measured Endmember Spectrum and Multi-spectrum Image of TM. Opto-Electron. Eng. 2012, 39, 62–70. 37. Shao, Z.; Cai, J. Remote sensing image fusion with deep convolutional neural network. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2018, 11, 1656–1669. [CrossRef] 38. Said, N.; Henning, B.; Joachim, H. Digital Mapping of Soil Properties Using Multivariate Statistical Analysis and ASTER Data in an Arid Region. Remote Sens. 2015, 7, 1181–1205. 39. Wang, X.; Zhang, F.; Ding, J.; Latif, A. Johnson VC. Estimation of soil salt content (SSC) in the Ebinur Lake Wetland National Nature Reserve (ELWNNR), Northwest China, based on a Bootstrap-BP neural network model and optimal spectral indices. Sci. Total Environ. 2018, 615, 918–930. [CrossRef][PubMed] 40. Tilley, D.R.; Ahmed, M.; Son, J.H.; Badrinarayanan, H. Hyperspectral Reflectance Response of Freshwater Macrophytes to Salinity in a Brackish Subtropical Marsh. J. Environ. Qual. 2007, 36, 780–789. [CrossRef][PubMed] 41. Chen, H.Y.; Zhao, G.X.; Chen, J.C.; Wang, R.Y.; Gao, M.X. Remote sensing inversion of saline soil salinity in the mouth of the Yellow River based on improved vegetation index. Trans. Chin. Soc. Agric. Eng. 2015, 31, 107–114. 42. Dong, J.J.; Niu, J.M.; Zhang, Q.; KangSa, R.L.; Han, F. Remote sensing yield estimation of typical grassland based on multi-source satellite data. Chin. J. Grassl. 2013, 35, 64–69. 43. Xu, L.; Wang, Q. Retrieval of soil water content in saline soils from emitted thermal infrared spectra using partial linear squares regression. Remote Sens. 2015, 7, 14646–14662. [CrossRef] 44. Bao, S.D. Soil Agrochemical Analysis; China Agriculture Press: Beijing, China, 2000. 45. Li, X.J.; Xin, X.Z.; Jiang, T. Spatial downscaling research of satellite land surface temperature based on spectral normalization index. Acta Geod. Cartogr. Sin. 2017, 46, 353–361. 46. Wu, W.C. The Generalized Difference Vegetation Index (GDVI) for Dryland Characterization. Remote Sens. 2014, 6, 1211–1233. [CrossRef] 47. Scudiero, E.; Skaggs, T.H.; Corwin, D.L. Regional-scale soil salinity assessment using Landsat ETM + canopy reflectance. Remote Sens. Environ. 2015, 169, 335–343. [CrossRef] 48. Fernández-Buces, N.; Siebe, C.; Cram, S.; Palacio, J.L. Mapping soil salinity using a combined spectral response index for bare soil and vegetation: A case study in the former lake Texcoco, Mexico. J. Arid. Environ. 2006, 65, 644–667. [CrossRef] 49. Qi, G.H.; Zhao, G.X.; Xi, X. Soil salinity inversion of winter wheat areas based on satellite-unmanned aerial vehicle-ground collaborative system in coastal of the Yellow River Delta. Sensors 2020, 20, 6521. [CrossRef][PubMed] 50. Francos, N.; Romano, N.; Nasta, P.; Zeng, Y.; Szabó, B.; Manfreda, S.; Ciraolo, G.; Mészáros, J.; Zhuang, R.; Su, B.; et al. Mapping Water Infiltration Rate Using Ground and UAV Hyperspectral Data: A Case Study of Alento, Italy. Remote Sens. 2021, 13, 2606–2636. [CrossRef] 51. Youme, O.; Bayet, T.; Dembele, J.; Cambier, C. Deep Learning and Remote Sensing: Detection of Dumping Waste Using UAV. Procedia Comput. Sci. 2021, 185, 361–369. [CrossRef] 52. Azarang, A.; Kehtarnavaz, N. Image fusion in remote sensing by multi-objective deep learning. Int. J. Remote Sen. 2020, 41, 9507–9524. [CrossRef] 53. Chen, L.; Liu, C.; Chang, F.; Li, S.; Nie, Z. Adaptive multi-level feature fusion and attention-based network for arbitrary-oriented object detection in remote sensing imagery. Neurocomputing 2021, 451, 67–80. [CrossRef] 54. Wang, J.; Huang, B.; Zhang, H.; Ma, P. Sentinel-2A Image Fusion Using a Machine Learning Approach. IEEE Geosci. Remote 2019, 57, 9589–9601. [CrossRef] 55. Li, J.; Guo, X.; Lu, G.; Zhang, B.; Xu, Y.; Wu, F.; Zhang, D. DRPL: Deep Regression Pair Learning For Multi-Focus Image Fusion. IEEE Trans Image Process 2020, 29, 4816–4831. [CrossRef] 56. Edoardo, A.C.; Costantini, G. Beyond the concept of dominant soil: Preserving pedodiversity in upscaling soil maps. Geoderma 2016, 271, 243–253. [CrossRef] 57. Zhang, X.; Zuo, W.; Zhao, S.; Jiang, L.; Chen, L.; Zhu, Y. Uncertainty in Upscaling In Situ Soil Moisture Observations to Multiscale Pixel Estimations with Kriging at the Field Level. ISPRS In. J. Geo-Inf. 2018, 7, 33. [CrossRef] 58. Cui, K.; Zhao, G.X.; Wang, Z.R.; Xi, X.; Gao, P.P.; Qi, G.H. Multi-scale spatial variability of soil salinity in typical fields of the Yellow River Delta in summer. Chin. J. Appl. Ecol. 2020, 31, 1451–1458. 59. Hu, J. Estimation of Soil Salinity in Arid Area Based on Multi-Source Remote Sensing. Master’s Thesis, Zhejiang University, Hangzhou, China, 2019. 60. Chen, W.J. Soil Salinity Retrieval and Dynamic Analysis Based Onspectral Intercalibration of Multi-Sensor Data. Master’s Thesis, Southeast University, Nanjing, China, 2018. 61. Chen, J.Y.; Wang, X.T.; Zhang, Z.T.; Han, J.; Yao, Z.H.; Wei, G.F. Soil Salinization Monitoring Method Based on UAV-Satellite Remote Sensing Scale-up. Trans. Chin. Soc. Agric. Mach. 2019, 50, 161–169.